KR20180106913A - Active filter - Google Patents

Abstract

The present invention discovers an electromagnetic interference (EMI) filter overcoming a problem of the latest technology. Disclosed is a power system for a vehicle. The power system comprises: a battery; a charging interface for receiving an external power for charging the battery; a network connecting the battery and the charging interface, wherein at least a part of the network is a DC network; and an active EMI filter between the battery and the charging interface in the DC network.

Description

Active filter {ACTIVE FILTER}

The present invention relates to active filters, and more particularly to active filters for DC networks and / or electric vehicles.

Electric vehicles have different power electronics conversion stages. The coexistence of such a large number of control and power signals in the limited space of the propulsion chamber of an electric vehicle poses a serious threat to the generation and propagation of electromagnetic interference (EMI). The use of shielded cables in powertrain systems with some filtering stages has become a standard method. However, if the power electronics transducer has undergone tremendous development over the past decade, it can not be said to be the same for passive components. The main concern is in magnetic components, and inductors and common mode chokes are widely used today to filter EMI; They are bulky and sometimes not suitable for mass production, and in fact high-level currents in electric vehicle powertrains have high cross-section cable or bus-bars that are understandable when they must be wound around the magnetic core Thereby increasing the need for use. Other technical issues are performance requirements, saturation current, and Curie temperature.

US6898092 suggests using a feedforward active filter for EMI. However, this document does not disclose how an active filter can be realized for high currents appearing in the powertrain of an electric vehicle. In addition, no method has been disclosed in which the active filter can realize a large filtering bandwidth of an EMI filter between 150 kHz and 30 MHz. In particular, for high frequencies, a very small delay in an already active circuit can lead to the fact that the elimination noise does not eliminate the power line noise, but increases the noise.

It is an object of the present invention to find an EMI filter that overcomes the problems of the state of the art.

This objective is achieved by using an active EMI filter in the network of electric vehicles.

Active EMI filters have advantages over passive EMI filters that do not scale with increasing currents in size and weight.

This object is also achieved by an active filter having a first power conductor, a second power conductor and an active circuit, the active circuit comprising: a sensing section for sensing noise in the first and / or second power conductors, And an injection section for injecting the elimination noise from the gain section into the first power conductor and / or the second power conductor.

Another embodiment of the present invention is described next.

In one embodiment, the injection section includes a coupling capacitance configured to inject rejection noise into the first power conductor and the second power conductor. Preferably, the coupling capacitance comprises a plurality of parallel capacitors having different capacitance values. This has the advantage that the coupling capacitance, thus the bandwidth characteristic of the injection section, is widened.

In one embodiment, the sensing section includes a current transformer sensing a noise current in the first power conductor and / or the second power conductor, and the injection section is configured to inject a rejection noise current into the first power conductor and the second power conductor Coupling capacitance, and the sensing section and the injection section are arranged in a feedback configuration. The bandwidth and performance of an active filter depends primarily on the electrical parameters of the device that generate the network and EMI noise. It has been found that the described topology and feedback configuration has proven to be the best solution for electric vehicle applications; It is more stable and immune to environmental changes such as temperature and longevity.

In one embodiment, the sensing section includes a current transformer having a core that inductively couples the first power conductor, the second power conductor, and the auxiliary conductor of the sensing circuit, and the auxiliary conductor is coupled to the gain section.

In one embodiment, the core material has a constant permeability between 150 kHz and 30 MHz, preferably between 150 kHz and 100 MHz. Although most applications use nanocrystalline core materials to achieve maximum permeability, they are not constant over the stated frequency range and are therefore not well suited for this application.

In one embodiment, the core material comprises MgZn. This material has proved to be the best fit to achieve this wide bandwidth.

In one embodiment, the auxiliary conductor includes a burden resistor or a transimpedance amplifier that is parallel to the current transformer to convert the measured noise current into a noise voltage.

In one embodiment, the auxiliary conductor comprises less than three windings around the core, preferably less than two windings. In one preferred embodiment, the auxiliary conductor comprises one winding or half a winding around the core. This has the advantage that the current transducer introduces low parasitic inductance and capacitance in an active circuit which reduces the phase shift produced by the active circuit.

In one embodiment, the coils of the auxiliary conductor may be formed with (solid copper) wire or through a PCB trace.

In one embodiment, each of the first and second conductors has a half turn of the core, i.e., the conductors are fed through the core without being wrapped around the core. This has the advantage that a conductor or bus bar with their large cross-sectional area need not be wound around the core for large currents. This simplifies the configuration of the filter.

In one embodiment, the first conductor is a first bus baud, the second conductor is a second bus bar, and the core is a ring core surrounding the first bus bar and the second bus bar. This has the advantage that bus bars with their large cross-sectional area need not be wound around the core for large currents. This simplifies the configuration of the filter. Preferably, the printed circuit board is arranged between the first bus bar and the second bus bar at least in the opening of the ring core. Similarly, the two conductors can be electrically isolated from each other despite their proximity.

In one embodiment, the printed circuit board (PCB) comprises an active circuit, and the first bus bar and / or the second bus bar are connected directly to the PCB for connection to the injection section. This has the advantage that an active circuit with its small current can be implemented on the PCB, while the bus bar can be directly connected to the PCB. This fastens and simplifies the fabrication of the filter.

In one embodiment, the first and second bus bars are connected to the PCB by direct soldering, screwing, or smt insert soldered and / or screwed to the bus bar and PCB.

In one embodiment, the current converter is a compensated (DC) current. This has the advantage that the flux from the DC current flowing in the two power conductors are arranged in opposite directions to each other and eliminated from each other. There is therefore no risk of saturation of the core material.

In one embodiment, the magnetic flux due to EMI currents flowing through the power conductors are combined with each other, thus realizing the sensing function of the current transducer.

In one embodiment, the printed circuit board comprises an active circuit, the core is a ring core, the PCB includes protrusions between two recesses and two recesses, Are accommodated in the two recesses so as to extend through the opening formed by the ring core. This arrangement allows a robust array of core and active filters. In addition, it is possible to realize the auxiliary conductor winding on the PCB.

In one embodiment, the active circuit is connected to the same ground of the system in which the filter is installed; Thus, the elimination noise current injected into the first power conductor and the second power conductor can flow back to the active circuit through the ground connection.

In one embodiment, the ground connection is a specific and essential feature of the active filter.

In one embodiment, the gain section comprises an operational amplifier. Operational amplifiers have been used due to their broadband performance and thermal stability.

In one embodiment, the EMI filter has a bandwidth of at least 150 kHz to 30 MHz.

In one embodiment, the EMI filter is for a DC network.

The described embodiments can be combined.

Although the active filter has been described for an automotive application or vehicle, it is also possible to apply this active filter for any DC application, such as other applications, especially solar cell energy applications. Although active filters have been described for filtering EMI, it is also possible to apply this filter for different frequencies.

The invention will be better understood with the aid of the description of an embodiment given by way of example and illustrated by the figures.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a view showing the principle of operation of an active filter; Fig.
2 is a schematic diagram of one embodiment of an active filter;
3 is a schematic diagram of four potential topologies for an active filter;
Figure 4 illustrates an embodiment of a circuit diagram of one embodiment of an active filter;
Figure 5 shows the impedance of different capacitors across frequency.
6 is a three-dimensional view of an embodiment of a mechanical configuration of an active filter in a mounted state.
Figure 7 is a three-dimensional view of an embodiment of the mechanical structure of Figure 4 in an unlocked state.
Figure 8 is a first side view of an embodiment of the mechanical structure of Figure 4;
Figure 9 is a second side view of an embodiment of the mechanical configuration of Figure 4;
Figure 10 is a third side view of an embodiment of the mechanical configuration of Figure 4;
11 is a diagram illustrating application of an active EMI filter to a power system of an electrically driven vehicle;

Fig. 1 shows the principle of operation of the present active filter. The DC network 21 has a first power conductor 11 and a second power conductor 12 for conducting the DC main current. The device 20 connected to the network through the first and second power conductors 11 and 12 generates noise, especially EMI. Noise flows on the first and second conductors 11 and 12 of the network 21 as shown by the solid line arrows and flows back to the device 20 through the ground connection. The active filter 22 is now installed between the device 20 and the network 21 to remove noise from the device 20 and to generate reject noise and to output reject noise to the first and second power conductors 11, . As a result, the noise is removed by the elimination noise between the active filter 22 and the network 21. The elimination noise flows from the active filter 22 to the network 21 and then back to the active filter 22 through the ground connection.

Preferably, the DC network 21 is configured for a current of at least 50 A, preferably at least 75 A, more preferably at least 100 A. Preferably, the voltage of the DC network 21 is greater than or equal to 2V. Preferably, the nominal voltage of DC network 21 is less than 800V. Preferably, the nominal voltage is 12V.

The active filter 22 is configured for such a DC network 21. The active filter 22 is preferably an EMI filter. The active filter 22 preferably has a filtering bandwidth between at least 150 kHz and 10 MHz, preferably between at least 150 kHz and 30 MHz. A filter having a bandwidth in the frequency range means that the filter reduces all noise frequencies in this frequency range in the first and second conductors 11 and 12 to attenuation greater than 0 dB. Preferably, the filter reduces all noise frequencies in this frequency range at the first and second conductors 11 and 12 to attenuation greater than 10 dB. The input of the active filter 22 is the side of the filter from which the noise comes, i.e. the side of the device 20 in FIG. The output of the active filter 22 is the side of the filter through which the noise flows, i. E. The side of the network 21 in Fig. The active filter 22 is preferably configured to filter common mode noise.

Figure 2 shows a schematic view of the active filter 22. The active filter 22 includes a first power conductor 11, a second power conductor 12, a sensing section 23, a gain section 24 and an injection section 25.

The first power conductor 11 and the second power conductor 12 of the active filter 22 are configured to conduct the main current between the device 20 and the network 21. The first power conductor 11 and the second power conductor 12 are configured for the above-mentioned current and / or voltage. Preferably, the first power conductor 11 and the second power conductor 12 have a cross-sectional area of at least 10 square millimeters. Preferably, the first power conductor 11 and the second power conductor 12 are busbars. Preferably, the first power conductor 11 and the second power conductor 12 are connected to the first power conductor 11 and the second power conductor 12 to connect the device 20 and the network 21, And has terminals on one or both sides. However, it is also possible to integrate an active filter into the device 20 or to another device or network 21. [

The sensing section 23 is configured to sense the noise of the first and second power conductors 11, 12. The points of the first and / or second power conductors 11 and / or 12, where the sensing section 23 senses the noise, are called at the next sensing point. The gain section 24 is configured to generate reject noise based on the sensed noise. The injection section 25 is configured to inject the removal noise into the first and second power conductors 11 and / or 12 so that the noise is removed (or at least reduced) by the noise removal noise. The points of the first and / or second power conductors 11 and / or 12 for which the injection section 25 injects the elimination noise are called at the next injection point.

There are many design options for providing active filters.

First, the active filter 22 may be arranged in a feedback or feedforward direction. The feedback direction means that the injection point is arranged in front of the detection point in the noise direction or upstream. That is, the feedback direction means that the injection point is arranged between the input part of the active filter 22 and the sensing point. Feed forward direction means that the injection point is arranged in the direction of the noise behind or downstream of the sensing point. That is, the feedforward direction means that the injection point is arranged between the output of the active filter 22 and the sensing point.

Second, there are four topologies for the active filter 22 as shown in FIG. The topology (a) measures the noise current and injects the elimination noise voltage. The topology (b) measures the noise current and injects the reject noise current. The topology (c) measures the noise voltage and injects the elimination noise current. The topology (d) measures the noise voltage and injects the elimination noise voltage. All four topologies of FIG. 3 are shown in the feedback direction. The topologies ((b) and (d)) may also be arranged in the feedforward direction. In the past, the parameters of the system, particularly the noise device 20 and the impedance of the network 21, have not been considered for selecting an active filter design. However, each of the performance of these six topologies, as well as the bandwidth and the parameters of the filter, the active filter 22, the load on the bandwidth of the (Z _L) (network 21 in Fig. 1) and the noise source (Z _S) ( Lt; RTI ID = 0.0 > 1) < / RTI > As a result, the choice of each filter design is highly application dependent. It has been found that the topology (b) of the feedback direction for an EMI filter, preferably with a wide bandwidth between at least 150 kHz and 10 MHz, preferably even between at least 150 kHz and 30 MHz, is particularly high performance for DC networks at vehicle verification. However, it is also possible to use one of the other five active filter designs described.

4 shows an embodiment of the circuit of the active filter 22. One preferred embodiment of the sensing section 23, the gain section 24 and the injection section 25 is described. The initiation of the circuitry of each section 23, 24 and 25 may also be combined with other implementations of other sections, although the next disclosure of the circuitry of each section 23, 24 and 25 is particularly beneficial in their combination.

In one embodiment, the sensing section 23 is configured to measure the noise current in the first power conductor 11 and in the second power conductor 12. The sensing section 23 preferably includes a current transformer for measuring the noise current in the frequency bandwidth of the active filter 22 current in the first power conductor 11 and in the second power conductor 12. The current transformer is used to inductively couple the first and second power conductors 11 and 12 with the auxiliary conductors to induce a noise current of the first power conductors 11 and of the second power conductors 12 in the auxiliary conductors And a core (14). Preferably, the core 14 couples the first and second power conductors 11 and 12 to produce a magnetic field in the core 14 where the same magnitude of current in the opposite direction removes (compensated DC current) from each other. The common mode current flowing in the same direction on the first and second power conductors 11 and 12 is derived in the auxiliary winding (in the frequency bandwidth of the active filter 22). Thus, preferably, the current transformer is a current compensated common mode current transformer. It has been found that one bottleneck in the bandwidth of the active filter 22 was the material of the core (and the number of primary and secondary windings of the current transducer). The primarily used nanocrystal core material can have high permeability, but its permeability depends on the large frequency bandwidth. It has been shown that the performance of the active filter 22 is significantly improved even though the core material with substantially constant permeability in the bandwidth of the active filter 22 is much lower in permeability. An example of such a core material is MgZn ferrite. Considering the high frequencies up to 30 MHz to be filtered, the small delay produced by the sensing section 23, the gain section 24 and the injection section 25 results in noise rather than eliminating it by injection of rejection noise. Can be amplified. Thus, the delay produced by the sensing section 23, the gain section 24 and the injection section 25 should be kept as low as possible, preferably below 4 nanoseconds. Another measure to achieve this is to use one or more windings around the core 14 in less than three windings, preferably less than two windings, preferably auxiliary conductor 27. [ This reduces the parasitic inductance and capacitance and thus reduces the delay caused by the current transducer. The winding direction of the auxiliary conductor 27 around the core 14 is preferably such that the induced or sensed current is opposite to the noise current in the first and second power conductors 11 and 12 , And the gain section 24 just amplifies the sensed noise without any inversion of the sensed noise (multiplies by -1). The inversion of the sensed noise can also be effectively provided by the OpAmp stage. The sensing section 23 preferably includes a burden resistor 28 or a transimpedance amplifier for delivering the sensed noise current to the sensed noise voltage so that the sensing section 23 senses the noise voltage sensed by the gain section 24 . The burden resistor 28 is connected in parallel to the secondary winding of the current transformer. Although other circuit arrangements are possible, the arrangement of burden resistors in parallel to the secondary windings of the current transducer proved to be particularly high performance. Preferably, one side of the burden resistor 28 and / or one side of the windings of the auxiliary conductor 27 are connected to ground.

The gain section 24 is configured to generate reject noise based on the sensed noise. Preferably, the gain section 24 is configured to amplify the sensed noise in the sensing section 23. Preferably, the gain section 24 receives the noise voltage from the sensing section 23 through the burden resistor 28, for example as a noise voltage. The gain section 24 preferably includes an operational amplifier 36. The two inputs of the operational amplifier are preferably connected to the noise voltage from the sensing section 24. [ Preferably, the first input (preferably positive input) of the operational amplifier 36 is connected to the first terminal of the windings of the burden resistor 28 and / or the auxiliary conductor 27 of the sensing section 23 do. Preferably, the second input (preferably negative input) of the operational amplifier is connected to ground and / or to the second terminal of the windings of the burden resistor 28 and / or the auxiliary conductor 27 of the sensing section 23, (Via resistor 29). Preferably, the operational amplifier 36 is operated via a closed loop feedback control, that is, the second terminal is connected to the output of the operational amplifier 36 via the resistor 30. The gain section 24 may include another amplifier stage. Preferably, the output of the operational amplifier 36 is supplied to the first (preferably positive) input of the second operational amplifier 37 via the resistor 38. The second (preferably negative) input of the second operational amplifier 37 is connected to the output of the second operational amplifier 37 (via the resistor 33). The first input of the second operational amplifier 37 is preferably connected to ground (via resistor 32). The output of the first or second operational amplifier 36 or 37 is connected to the input of the injection section 25 (via resistor 34). The circuit described is just one example of how the gain section 24 can be realized. However, this embodiment shows a particularly high gain section 24 in the bandwidth of the filter. Other suitable OpAmp configurations for the gain section 24 are "2 OpAmp instrumentation amplifier" and "3 OpAmp instrumentation amplifier ".

In one embodiment, the injection section 25 is configured to inject the removed noise current from the gain section 24 into the first and second power conductors 11 and 12 such that the noise current is removed Or at least reduced). Preferably, the coupling capacitance 13 is used to inject the elimination noise into the first and second power conductors 11 and 12. Preferably, the elimination noise is injected into the first power conductor 13 through the first coupling capacitance 13 and into the second power conductor 12 through the second coupling capacitance 13. It has been found that the bandwidth of the coupling capacitance 13 is important for the performance of the active filter 22. The impedance of the ideal capacitor decreases with respect to the high frequency. However, due to the parasitic inductance in the actual capacitor, the impedance of the capacitor begins to increase again after the series resonance frequency of the capacitance and parasitic inductance. Figure 5 shows the impedance characteristics of other capacitors over frequency. In one preferred embodiment, each coupling capacitance 13 comprises a plurality of parallel capacitors having different capacitance values. As a result, the elimination noise for each frequency can always select a path with a capacitor that provides the lowest impedance for this frequency (the total impedance of the parallel capacitor is dominant for each frequency with the smallest impedance) . Each of the parallel capacitors of the coupling capacitance 13 is connected to the output of the gain section 24 on one side and is connected to the first point on the other side between the input point of the active filter 12 and the sensing section 23, Or the second power conductor (11 or 12).

Through the preferred embodiment of the active filter 22, active noise attenuation of 60 dB at 100 kHz, 40 dB at 1 MHz and 20 dB at 10 MHz has been achieved. This active filter 22 is a high performance first active filter over this large bandwidth. It is also a first active filter 22 whose bandwidth performance is optimized for each of the sensing, gain and injection sections.

A preferred application of this active filter 22 is in electric motor bikes, cars and trucks as examples of electric vehicles. In particular, the DC network having the first power conductor 11 and the second power conductor 12 between the battery 40 and the charging interface 41 can be used as such an EMI active filter for EMI removal, (22). The charging interface 41 often includes a power converter that converts power from an external network (often AC) to DC. These power converters are often the source of EMI. Battery management is another source of EMI. In particular, the input and / or output of the battery 40 and / or another potential component of the charge interface 41 and / or similar to the power converter can be connected to this EMI active filter 22, respectively.

Figures 6, 8, 9 and 10 illustrate the preferred structural design of the active filter 2 described. Fig. 6 shows a three-dimensional view, while Figs. 8, 9 and 10 show three sides of the filter 22. Fig. The gain section 24 and the injection section 25 are preferably implemented in the PCB 15. The core 14 is arranged around the first power conductor 11 and the second power conductor 12 in a feed-through arrangement. The core 14 is preferably a ring core, preferably a closed ring core with no air gap, and the first power conductor 11 and the second power conductor 12 are connected to each other through an opening in the ring core 14 Or also having a half-turn). The ring core 14 is not limited to a circular ring shape. Other ring cross-sectional shapes such as ellipsoid, secondary, and rectangle are also possible. Preferably, the first power conductor 11 and the second power conductor 12 are busbars. Preferably, the bus bar is directly connected to the PCB to be conductively connected to the injection section 25 of the PCB 15, for example. The bus bar can be connected to the PCB by soldering, screwing or PCB inserts. Preferably, the PCB 15 is arranged between the two bus bars 11, 12 to separate them from each other. Thus, the PCB 15 not only carries the circuitry of the gain section 24 and the injection section 25, but also has some other function. This mechanically holds the bus bars 11 and 12, which creates and / or separates the electrical connections between the bus bars 11 and 12 and the injection section 25 from one another. As shown in Fig. 8, the bus bars 11, 12 are formed or meandered so that they have different spacing in different areas. In the region of the core 14 and / or the area of the mechanical and electrical connections between the bus bars 11, 12 and the PCB 15, the bus bars 11, 12 are smaller than the ends of the bus bars 11, Where the active filter 22 provides the input and output terminals of the active filter 22. Preferably, the bus bars 11, 12 have a rectangular cross section with two flat sides and two small sides. The bus bars 11, 12 are arranged on the PCB 15 with flat sides. The PCB 15 includes another two recesses for receiving the ring core 14. 7 shows how the ring core 14 having two bus bars 11, 12 penetrating the opening of the core 14 can be inserted into the two recesses. The PCB 15 forms a protrusion 15.1 through the opening of the core 14 between the two bus bars 11, 12 between the two recesses. Thus, the projections 15.1 of the PCB 15 separate them from each other despite the close arrangement of the two bus bars 11, 12. The protrusion 15.1 also allows the winding of the auxiliary conductor 27 to be realized as a conductor on the PCB 15 (not shown). In this case, the auxiliary conductor 27 is connected to the rest of the PCB 15 by a bridge or wire through the recess from the projection 15.1. However, it is also possible to realize the winding of the auxiliary conductor 27 with a wire. The active filter 22 further includes a housing (not shown) covering the PCB 15 and the core 14 so that only the four ends of the two bus bars 11, 12 protrude from the filter housing.

Claims (17)

1. An EMI filter for a DC network comprising a first power conductor (11), a second power conductor (12) and an active circuit,
The active circuit comprises:
A sensing section 23 for sensing noise in the first and second power conductors 11 and 12,
A gain section (24) for generating an elimination noise based on the noise sensed in the sensing section (23), and
And an injection section (25) for injecting said elimination noise from said gain section into said first power conductor (11) or said second power conductor (12). The method according to claim 1,
The injection section (25) comprises a first coupling capacitance (13) configured to inject the removal noise into the first power conductor (11), the first coupling capacitance (13) comprising a plurality of An EMI filter comprising a parallel capacitor. 3. The method of claim 2,
The injection section 25 comprises a second coupling capacitance 13 configured to inject the elimination noise into the second power conductor 12 and the second coupling capacitance 13 comprises a plurality of An EMI filter comprising a parallel capacitor. The method of claim 3,
The gain section includes an output point for outputting the elimination noise and the first coupling capacitance is connected to the first power conductor 11 through a first terminal and the gain section 24 through a second terminal. And the second coupling capacitance 13 is connected to the second power conductor 12 through a first terminal and to the output point of the gain section 24 via a second terminal, EMI filter. The method according to claim 1,
Wherein the sensing section comprises a current transformer having a core (14) for inductively coupling a first power conductor (11), a second power conductor (12) and an auxiliary conductor (27) of the active circuit, (27) is connected to the gain section (24). 6. The method of claim 5,
Wherein the material of the core (14) has a constant permeability between 150 kHz and 30 MHz. 6. The method of claim 5,
Wherein the material of the core comprises MgZn. 6. The method of claim 5,
Wherein the sensing section (23) comprises a burden resistor (28) in parallel with the current transducer to convert the measured noise current to a noise voltage. 6. The method of claim 5,
The auxiliary conductor (27) comprises less than three windings around the core (14), preferably less than two windings. 6. The method of claim 5,
Wherein the first power conductor 11 is a first bus bar and the second power conductor 12 is a second bus bar and the core is a ring that surrounds the first bus bar and the second bus bar, Core, EMI filter. 11. The method of claim 10,
A printed circuit board (15) is arranged between the first bus bar and the second bus bar at least in the opening of the ring core (14). 6. The method of claim 5,
The printed circuit board 15 includes the gain section 24 and the injection section 25, wherein the first power conductor 11 is a first bus-powered and the second power conductor 12 is a second bus- Wherein the first bus bar and / or the second bus bar are directly connected on the PCB for electrical connection with the injection section (25). 6. The method of claim 5,
The printed circuit board includes the gain section (24) and the injection section (25), the core (14) is a ring core, and the PCB includes two recesses and projections between the two recesses And the ring core is received in the two recesses so that the protrusion (15.1) extends through an opening formed by the ring core. The method according to claim 1,
Wherein the active circuit is connected to ground so that the reject noise current injected into the first power conductor (11) and the second power conductor (12) can flow back to the active circuit through the ground connection. The method according to claim 1,
Wherein the active circuit's rejection noise is configured to reduce noise in a bandwidth between 150 kHz and 30 MHz. The method according to claim 1,
The sensing section 23 comprises a current transformer for sensing the noise current in the first power conductor 11 and the second power conductor 12 and the gain section 24 is connected in the sense section 23 And the injection section (25) is configured to generate the elimination noise from the gain section via the coupling capacitance (13) to the first power conductor (11) and the second power conductor Is configured for injecting into a second power conductor (12), said sensing section (23) and said injection section (25) being arranged in a feedback direction. A power system for an electric powered vehicle comprising:
battery;
A charging interface for receiving external power to charge the battery;
A network connecting the battery and the charging interface, wherein at least a portion of the network is a DC network;
And an EMI filter (22) according to any one of claims 1 to 16 between the battery and the charging interface in the DC network.

Images